DOI: https://doi.org/10.21203/rs.3.rs-93848/v1
Background: Stressors introduced to adolescents by COVID-19 social distancing measures may cause mental health problems to (re)surface. We studied depression, anxiety and stress among adolescents experiencing lockdown.
Methods: From May-June 2020, secondary school students were enrolled in an online cross-sectional survey through social media. We assessed presence and severity of depression (PHQ-9), anxiety (GAD-7) and stress (PSS-10) in the last month, demographics, degree of social distancing, and other associated issues.
Results: Of 392 respondents (56.4% male, 43.1% female), mostly from Thailand (59.2%) and UK (26.5%). We identified depressive symptoms in 58.7%, anxiety in 40.3% and high levels of stress in 9.7%. We found, by multivariate analysis, significant associations between being female and depression and anxiety, being in late secondary school years and depression, and changes in patterns of substance use and anxiety and stress.
Conclusions: We propose that girl-centred mental health support platforms should be readily available and tailored to fit specific countries’ contexts. Schools must closely monitor and act upon any concerns which arise from their students and must also monitor mental health wellbeing as changes in academic routine due to COVID-19 could be drastic for some. Harm reduction services must adapt and utilise innovative telemedicine interventions, tailored towards adolescent users.
Due to the current 2019 novel coronavirus (SARS-CoV-2) pandemic, many governments have initiated unavoidable social distancing measures to attempt to slow the spread of disease, in the form of widespread lockdown of varying severity as well as other ways. This is not new: citywide lockdowns were previously implemented in countries such as Canada and China during the 2003 severe acute respiratory syndrome (SARS) outbreaks, and efforts were made to quarantine whole villages in West Africa during the 2014 Ebola outbreak (1). For adolescents, such lockdowns result in school closures, online learning, exam cancellations as well as being restricted socially in other ways. By the end of July 2020, over a billion enrolled learners were still out of “normal” education (2).
Adolescent mental health problems are becoming an increasingly apparent issue. Adolescence can prove to be a crucial time of mental development, plagued with higher risks of developing psychiatric disorders. Children and adolescents have been shown to be one of the most vulnerable groups with regards to mental health during extended isolation (3). Even after lockdown measures in the United Kingdom (UK) were lifted, the University College London COVID-19 Social Study reported that young people were one of the groups still experiencing high levels of depression and anxiety (4).
In general, lockdown and its element of social and physical distancing elicits a toll on mental health; the World Health Organization (WHO) affirms that social dysfunction will result in elevated prevalence of psychiatric illnesses (5). Hawryluck et al. showed that persons quarantined in Toronto, Canada, during the 2004 SARS outbreak displayed a high prevalence of depression (6). Jeong et al. reported higher rates of anxiety among individuals during their isolation period compared to 4–6 months after isolation due to the 2015 Middle East Respiratory Syndrome (MERS) outbreak in Korea (7). More recently, Huang and Zhao have identified a “major mental health burden” on the Chinese public amid the COVID-19 outbreak (8).
Since the pandemic’s inception, there have been some studies conducted which have begun to assess how COVID-19 social distancing and isolation can lead to adverse mental health effects (8, 9) and further consequences such as substance abuse (10–12). However, evidence and data on how large-scale public health measures affect adolescent mental health outcomes is scarce (13). Xie et al. report higher prevalence of depression and anxiety among primary school children in home confinement during a Chinese nationwide school closure (14). Zhou et al. describe similar findings with Chinese high school students (15). Despite this, there have not been any studies examining the disparities between mental health impact from country to country, and certain associated factors such as adolescent substance use.
Thailand and the UK are two countries which, despite similar population sizes, have experienced very different trajectories regarding COVID-19 outbreak. Thailand reported the first case of COVID-19 outside of China on 13 January 2020 (16), with the virus reaching the UK shortly after later in the month (17). In response, social distancing measures were implemented by both governments. On 21 March 2020, the Bangkok Metropolitan Administration authority declared widespread shutdown of various businesses. A national public state of emergency was declared on 25 March, with a general lockdown and social distancing requirements instituted on 26 March. In the UK, governmental response was initially in the form of guidance. As the situation escalated, legislation was enacted in the form of statutory instruments, which included implementing closures of schools, businesses and non-essential services, restrictions on movement and gatherings, and enforcement. A stay-at-home order came into effect on 26 March. Despite taking action at around similar times, the two countries currently find themselves in dissimilar situations. Since then, as of 7 September 2020, Thailand has had 3,445 COVID-19 cases, compared to 347,152 cases in the UK. Their death rates per million people differ substantially: 0.8 for Thailand, but 611 for the UK (making it the sixth-highest death rate per million people globally among major countries) (18). Thus, it would be interesting to examine the differences in adolescent mental health amid a worldwide pandemic, but in differing COVID-19 situations.
We therefore aimed to pilot assess the impact of social distancing on the mental health of secondary school students in Thailand and UK, including the degree and prevalence of depression, anxiety and stress, as well as how lockdown-specific associated factors can impact this. However, there will be no geographical boundaries due to the use of online survey.
We advertised a study survey link launched to students in Thailand and UK via snowball sampling on social media popular with youths, including Instagram, Snapchat, WhatsApp and LINE. The location of study had no physical space, and participants instead accessed the digital platform of Google Forms on an online system. Further distribution of the survey link was done by earlier participants via instant messaging as well as ‘story reposts’. In order to answer the questions, participants had to be enrolled in school from Year 8 (Grade 7) to Year 13 (Grade 12) and be able to understand, read and write (type) English.
Consent by action was obtained via a digitalised consent form, explaining the objectives and contents of the survey as well as potential risks and intended benefits. We designed the online survey layout such that no study procedures would occur prior to the participant giving informed consent. Despite most participants being under 18, parental consent was waived due to the anonymous nature of the survey.
The study was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University.
The cross-sectional survey on mental health was created using the online service Google Forms, and participants could access the survey and answer questions on it. The survey was open to responses from May to June 2020. Content included diagnostic instruments used to assess symptoms of depression, anxiety and stress, but also questions designed to identify general demographics, degree of social distancing and other associated factors. Open-ended questions were also implemented in order to allow participants to expand on certain issues or topics.
Prior to official deployment, 10 individuals were selected to test the questionnaire. These individuals were specifically selected so that feedback obtained later on was based on demographically diverse opinions.
Study measurements
The 9-item Patient Health Questionnaire-9 (PHQ-9) was used to assess depressive symptoms. PHQ-9 forms the depression module of the Patient Health Questionnaire (PHQ), which is the self-administered version of the PRIME-MD diagnostic tool developed by Pfizer. It has been validated for adolescent use (19). Participants scored their frequency of experience with each of the nine DSM-IV criteria over the last two weeks (e.g. poor appetite or overeating, feeling tired or having little energy, with their 4-point Likert-scale ratings representing frequencies (‘0’ is not at all, ‘1’ is several days, ‘2’ is more than half the days, ‘3’ is nearly every day). Scores were then totaled, with cut off points correlating to level of perceived depression (0–4 = none, 5–9 = mild, 10–14 = moderate, 15–19 = moderately severe, 20–27 = severe). Participants who scored above 4 were considered to exhibit depressive symptoms.
The 7-item Generalized Anxiety Disorder Scale (GAD-7) was used to assess anxiety symptoms. GAD-7 forms the anxiety module of the PHQ. The GAD-7 items correspond with DSM-IV criteria and use the same 4-point Likert-scale ratings as PHQ-9 to assess frequency of experience with criteria over the last two weeks, and thus presence and severity of anxiety. Participants rated each item, and scores were then totaled. Cut-off points correlating to level of perceived anxiety were: 0–4 = none, 5–9 = mild, 10–14 = moderate, 15–21 = severe. Participants who scored above 4 were considered to exhibit anxiety symptoms.
The Perceived Stress Scale (PSS-10) was used to assess participants’ perception of stress. Participants scored the frequency of ten different relevant items (e.g. feeling that ‘things were going your way’) in the last month using a 5-point Likert-scale system (0 = never, 1 = almost never, 2 = sometimes, 3 = fairly often, 4 = very often). Four positively stated items (namely questions 2, 4, 5 and 10) had their ratings reversed (0 = 4, 1 = 3, 2 = 2, 3 = 1 and 4 = 0) before ratings were totaled to score participants. Different score intervals correlated with different levels of perceived stress (0–13 = low, 14–26 = moderate, 27–40 = high).
Participants inputted general demographical information such as gender, age, school year, country currently residing in and country studying in. We also asked questions relevant to their lockdown experience, regarding topics such as substance use (previous and current uses as well as changes in patterns of use) and issues with online learning and exam cancellations (experience, and whether this was problematic). Participants were also able to report any previously diagnosed mental illnesses, and whether they experienced any change in this during lockdown.
For depression and anxiety, higher levels correspond to having a total PHQ-9/GAD-7 score of more than 4 (0–4 = symptoms not present, 4 + = symptoms present). For stress, higher levels correspond to having a total PSS-10 score of more than 26 (0–26 = low/moderate, 26 + = high).
Demographical and lockdown-related variables were compared between participants of differing levels of depression (Present/Not present), anxiety (Present/Not present) and perceived stress (‘low’/‘moderate to high’). Chi-square test was used for categorical variables, while Student’s t-test was used for continuous variables. Logistic (nominal) regression analysis was used to evaluate significant associations between predictive variables and presence of depression, anxiety and severity of perceived stress. Odds ratios (ORs) of predictive factors were reported, together with 95% confidence intervals (CIs). Statistical significance was set at p < 0.05. SPSS software (version 22) was used to perform statistical analysis.
Of 392 respondents (56.4% male, 43.1% female, 0.3% non-binary), mean (SD) age was 15.5 (1.7) years. Respondents reported currently living in Thailand (59.2%), the United Kingdom (26.5%), Hong Kong (3.3%), Singapore (1.8%), the United States (1.5%), Malaysia (1.2%), China (0.8%), Australia (0.5%), Belgium (0.5%), India (0.5%), Russia (0.5%), Saudi Arabia (0.5%), Italy (0.3%), Japan (0.3%), South Korea (0.3%), Luxembourg (0.3%), New Zealand (0.3%), Nigeria (0.3%), Turkey (0.3%) and Zambia (0.3%). Majority were in Year 12 (34.2%) and Year 10 (20.9%).
Since their physical school closure, 82.9% reported not having used public transport at all, while 52.3% had not talked face-to-face with someone not in their household. Around half (46.4%) found not being able to go out to shops problematic. Alcohol use was reported in 37.2% (11.0% increased usage, 14.0% reduced usage while 12.2% stayed the same). Use of cigarettes, e-cigarettes and vapes was reported in 17.6% (6.6% increased usage, 7.7% reduced usage while 3.3% stayed the same). Cannabis use was reported in 13.1% (3.6% increased usage, 6.4% reduced usage while 3.1% stayed the same). Having undergone mandatory self-quarantine was reported by 49.7%. 92.1% reported quarantining with their parents, 1.8% with family but without parents, and 5.5% with others.
Depression was identified in 58.7% (29.8% mild, 16.1% moderate, 7.9% moderately severe, 4.8% severe), and anxiety in 40.3% (22.2% mild, 10.7% moderate, 7.4% severe). Severities of stress ranged from low (35.7%) to moderate (54.6%) to high (9.7%). 21.4% reported previous depression (9.4% worsened, 8.7% unchanged, 3.3% improved). 35.5% reported previous anxiety (15.8% worsened, 13.3% unchanged, 6.4% improved). 45.2% reported previous stress (19.4% worsened, 16.1% unchanged, 9.7% improved). For statistical analysis, we categorised depression and anxiety into two groups, one indicating presence (those who scored mild and above) and another indicating absence of mental illness symptoms. Stress was categorised to form two groups: “High” and “Moderate/Low”.
Factors associated with depressive symptoms
Exhibiting depressive symptoms | No depressive symptoms | ||||||
---|---|---|---|---|---|---|---|
n | % | n | % | χ2 | df | p-values | |
UK School Year | |||||||
8/9/10 | 67 | 40.1 | 100 | 59.9 | 41.306 | 1 | < 0.001 |
11/12/13 | 163 | 72.4 | 62 | 27.6 | |||
Female | 108 | 63.9 | 61 | 36.1 | 4.015 | 2 | 0.134 |
Age (mean / SD) | (15.9) | (1.6) | (15.0) | (1.8) | t388= -5.274 | - | < 0.001 |
Country currently living in | |||||||
Thailand | 116 | 50.0 | 116 | 50.0 | 20.698 | 2 | < 0.001 |
UK | 77 | 74.0 | 27 | 26.0 | |||
Others | 37 | 71.2 | 154 | 28.8 | |||
Country currently studying in | |||||||
Thailand | 101 | 52.3 | 92 | 47.7 | 6.949 | 2 | 0.031 |
UK | 108 | 64.7 | 59 | 35.3 | |||
Others | 20 | 69.0 | 9 | 31.0 | |||
Area currently living in | |||||||
Urban | 138 | 56.1 | 108 | 43.9 | 4.521 | 2 | 0.104 |
Peri-urban | 49 | 57.6 | 36 | 42.4 | |||
Rural | 42 | 71.2 | 17 | 28.8 | |||
Being under mandatory self-quarantine | 128 | 65.6 | 67 | 34.4 | 7.769 | 1 | 0.005 |
Level of feeling that COVID-19 affects daily life | |||||||
Great effect | 140 | 64.5 | 77 | 35.5 | 10.066 | 2 | 0.007 |
Slight/little effect | 83 | 53.9 | 71 | 46.1 | |||
No effect | 7 | 33.3 | 14 | 66.7 | |||
Using public transport since the physical closure of school | 21 | 31.3 | 46 | 68.7 | 24.894 | 1 | < 0.001 |
Level of thinking that catching COVID-19 | |||||||
High chance | 7 | 33.3 | 14 | 66.7 | 6.006 | 2 | 0.050 |
Moderate chance | 84 | 61.3 | 53 | 38.7 | |||
Little/no chance | 139 | 59.4 | 95 | 40.6 | |||
Being worried about going outside | |||||||
Not at all | 83 | 64.8 | 45 | 35.2 | 18.622 | 2 | < 0.001 |
Slightly/a bit worried | 121 | 63.0 | 71 | 37.0 | |||
Very worried | 26 | 36.1 | 46 | 63.9 | |||
Experience problems with sleep by thinking about COVID-19 | 61 | 67.0 | 30 | 33.0 | 3.415 | 1 | 0.065 |
Online learning | |||||||
Experience with problems | 170 | 68.5 | 78 | 31.5 | 30.117 | 2 | < 0.001 |
Experience without any problems | 43 | 38.1 | 70 | 61.9 | |||
No experience | 16 | 53.3 | 14 | 46.7 | |||
Cannot meet friends regularly | |||||||
Experience with problems | 188 | 62.3 | 114 | 37.7 | 8.020 | 2 | 0.018 |
Experience without any problems | 35 | 50.7 | 34 | 49.3 | |||
No experience | 7 | 35.0 | 13 | 65.0 | |||
Cannot go out to eat | |||||||
Experience with problems | 142 | 64.3 | 79 | 35.7 | 9.753 | 2 | 0.008 |
Experience without any problems | 78 | 54.9 | 64 | 45.1 | |||
No experience | 10 | 35.7 | 18 | 64.3 | |||
Cannot go out to shop | |||||||
Experience with problems | 134 | 73.6 | 48 | 26.4 | 31.840 | 2 | < 0.001 |
Experience without any problems | 77 | 47.8 | 84 | 52.2 | |||
No experience | 19 | 39.6 | 29 | 60.4 | |||
Depression | |||||||
Has decreased or stayed the same | 38 | 80.9 | 9 | 19.1 | 30.057 | 2 | < 0.001 |
Has increased | 33 | 89.2 | 4 | 10.8 | |||
Never had it | 159 | 51.6 | 149 | 48.4 | |||
Anxiety | |||||||
Has decreased or stayed the same | 44 | 57.1 | 33 | 42.9 | 49.361 | 2 | < 0.001 |
Has increased | 61 | 98.4 | 1 | 1.6 | |||
Never had it | 125 | 49.4 | 128 | 50.6 | |||
Stress | |||||||
Has decreased or stayed the same | 65 | 64.4 | 36 | 35.6 | 59.013 | 2 | < 0.001 |
Has increased | 71 | 93.4 | 5 | 6.6 | |||
Never had it | 94 | 43.7 | 121 | 56.3 | |||
Alcohol | |||||||
Never used | 123 | 50.4 | 121 | 49.6 | 19.009 | 2 | < 0.001 |
Reduced | 38 | 69.1 | 17 | 30.9 | |||
Used the same or increased | 68 | 74.7 | 23 | 25.3 | |||
Cigarettes, e-cigarettes, vapes | |||||||
Never used | 171 | 52.9 | 152 | 47.1 | 25.531 | 2 | < 0.001 |
Reduced | 24 | 80.0 | 6 | 20.0 | |||
Used the same or increased | 35 | 89.7 | 4 | 10.3 | |||
Cannabis | |||||||
Never used | 182 | 53.5 | 158 | 46.5 | 30.254 | 2 | < 0.001 |
Reduced | 23 | 92.0 | 2 | 8.0 | |||
Used the same or increased | 25 | 96.2 | 1 | 3.8 | |||
Other drugs | |||||||
Never used | 196 | 55.2 | 159 | 44.8 | 17.901 | 2 | < 0.001 |
Reduced | 22 | 91.7 | 2 | 8.3 | |||
Used the same or increased | 11 | 91.7 | 1 | 8.3 | |||
Univariate analysis was conducted using chi-square test and t-test. p < 0.05 was considered significant. |
df, degrees of freedom; SD, standard deviation; UK, United Kingdom.
Higher levels of depression were found in Years 11 and higher (72.4% vs 40.1%, p < 0.001), residents of the UK and other countries versus Thailand (74.0%, 71.2% vs 50.0%, p < 0.001), those who studied in the UK and other countries versus Thailand, (69.0%, 64.7% vs 52.3%, p = 0.031) and older participants (M = 15.90, SD = 1.58 for depressed vs M = 15.01, SD = 1.755 for not depressed, p < 0.001) (Table 1).
Depression was higher in those who reported previous depression, anxiety, and stress (p < 0.001). Higher levels of depression were also reported in those who had undergone mandatory self-quarantine, felt COVID-19 had affected their daily life, reported not using public transport since school closure, believed that they would have low chance of contracting COVID-19 themselves, and had no worries about going outside (p < 0.05). In addition, those who perceived problems with the consequences of social distancing, including not being able to meet friends, inability to eat out, inability to go shopping and online learning, all displayed higher levels of depression (p < 0.05) (Table 1).
Those who reported use of alcohol, cigarettes/e-cigarettes/vapes, cannabis, and/or other drugs experienced higher levels of depression than those who never used them. Furthermore, those who experienced unchanged or increased use of alcohol, cigarettes/e-cigarettes/vapes and cannabis during lockdown experienced higher levels of depression than those who experienced reduced use of these substances (p < 0.001) (Table 2).
By multivariate logistic regression analysis, we found that older year groups (Year 11/12/13 compared to Years 8/9/10, OR = 2.255, 95% CI = 1.11–4.58), females (compared to males, OR = 2.46, 95% CI = 1.19–5.08) and participants not located in the UK or Thailand (compared to being in Thailand, OR = 3.07, 95% CI = 1.055–8.94) had significantly increased risk of depression (Table 2).
Variables | Exhibiting depressive symptoms | ||
---|---|---|---|
ORs | 95% CI | p-values | |
UK School Year | |||
11/12/13 | 2.255 | 1.11–4.58 | 0.024 |
8/9/10 | 1 | ||
Gender | |||
Female | 2.46 | 1.19–5.08 | 0.015 |
Male | 1 | ||
Depression | |||
Has decreased or stayed the same | 4.91 | 1.33–18.20 | 0.017 |
Has increased | 0.88 | 0.145–5.38 | 0.892 |
Never had it | 1 | ||
Anxiety | |||
Has decreased or stayed the same | 0.575 | 0.19–1.71 | 0.320 |
Has increased | 15.43 | 1.46–163.05 | 0.023 |
Never had it | 1 | ||
Stress | |||
Has decreased or stayed the same | 1.64 | 0.63–4.30 | 0.313 |
Has increased | 17.62 | 4.04–76.81 | 0.000 |
Never had it | 1 | ||
Country currently living in | |||
UK | 2.29 | 0.94–5.625 | 0.07 |
Others | 3.07 | 1.055–8.94 | 0.04 |
Thailand | 1 | ||
Cannabis | |||
Used the same or increased | 5.76 | 0.34–96.69 | 0.224 |
Reduced | 10.20 | 1.19–87.18 | 0.034 |
Never used | 1 | ||
Multivariate analysis was conducted using logistic nominal regression. p < 0.05 was considered significant. |
OR, odds ratio; CI, confidence interval; UK, United Kingdom.
Higher levels of depression were found to be significantly associated with an increase in levels of previous anxiety (OR = 15.43, 95% CI = 1.46-163.05), and previous stress (OR = 17.62, 95% CI = 4.04–76.81) during lockdown, compared to those who reported never having previous anxiety or stress. Those who reported decreased or equal levels of previous depression during lockdown still had higher levels of depression than those who never had previous depression before the lockdown (OR = 4.91, 95% CI = 1.33–18.20).
Compared to those who had never used cannabis, previous users who experienced decreased use during lockdown had higher levels of depression (OR = 10.20, 95% CI = 1.19–87.18).
Factors associated with anxiety symptoms
Variables | Exhibiting anxiety symptoms | No anxiety symptoms | |||||
---|---|---|---|---|---|---|---|
n | % | n | % | χ2 | df | p-values | |
UK School Year | 25.627 | 1 | < 0.001 | ||||
8/9/10 | 43 | 25.7 | 124 | 74.3 | |||
11/12/13 | 115 | 51.1 | 110 | 48.9 | |||
Female | 85 | 50.3 | 84 | 49.7 | 13.961 | 2 | 0.001 |
Age (mean/SD) | (15.2) | (1.8) | (15.2) | (1.6) | t388= -4.232 | - | 0.005 |
Country currently living in | |||||||
Thailand | 82 | 35.3 | 150 | 64.7 | 6.924 | 2 | 0.031 |
UK | 51 | 49.0 | 53 | 51.0 | |||
Others | 25 | 48.1 | 27 | 51.9 | |||
Country currently studying in | 2.014 | 2 | 0.365 | ||||
Thailand | 72 | 37.3 | 121 | 62.7 | |||
UK | 72 | 43.1 | 95 | 56.9 | |||
Others | 14 | 48.3 | 15 | 51.7 | |||
Area currently living in | |||||||
Urban | 94 | 38.2 | 152 | 61.8 | 4.371 | 2 | 0.112 |
Peri-urban | 32 | 37.6 | 53 | 62.4 | |||
Rural | 31 | 52.5 | 28 | 47.5 | |||
Being under mandatory self-quarantine? | 85 | 43.6 | 110 | 56.4 | 1.739 | 1 | 0.187 |
Level of feeling that COVID-19 affects daily life | |||||||
Great effect | 95 | 43.8 | 122 | 56.2 | 2.984 | 2 | 0.225 |
Slight/little effect | 6 | 28.6 | 15 | 71.4 | |||
No effect | 57 | 37.0 | 97 | 63.0 | |||
Using public transport since the physical closure of school | 9 | 13.4 | 58 | 86.6 | 24.256 | 1 | < 0.001 |
Level of thinking that catching COVID-19 | |||||||
High chance | 6 | 28.6 | 15 | 71.4 | 1.741 | 2 | 0.419 |
Moderate chance | 53 | 38.7 | 84 | 61.3 | |||
Little/no chance | 99 | 42.3 | 135 | 57.7 | |||
Being worried about going outside | |||||||
Not at all | 52 | 40.6 | 76 | 59.4 | 11.265 | 2 | 0.004 |
Slightly/a bit worried | 89 | 46.4 | 103 | 53.6 | |||
Very worried | 17 | 23.6 | 55 | 76.4 | |||
Problems with sleep by thinking about COVID-19? | 50 | 54.9 | 41 | 45.1 | 10.556 | 1 | 0.001 |
Online learning | |||||||
Experience with problems | 112 | 45.2 | 136 | 54.8 | 7.504 | 2 | 0.023 |
Experience without any problems | 34 | 30.1 | 79 | 69.9 | |||
No experience | 11 | 36.7 | 19 | 63.3 | |||
Cannot meet friends regularly | |||||||
Experience with problems | 129 | 42.7 | 173 | 57.3 | 3.546 | 2 | 0.170 |
Experience without any problems | 24 | 34.8 | 45 | 65.2 | |||
No experience | 5 | 25.0 | 15 | 75.0 | |||
Cannot go out to eat | |||||||
Experience with problems | 101 | 45.7 | 120 | 54.3 | 6.925 | 2 | 0.031 |
Experience without any problems | 50 | 35.2 | 92 | 64.8 | |||
No experience | 7 | 25.0 | 21 | 75.0 | |||
Cannot go out to shop | |||||||
Experience with problems | 91 | 50.0 | 91 | 50.0 | 13.646 | 2 | 0.001 |
Experience without any problems | 54 | 33.5 | 107 | 66.5 | |||
No experience | 13 | 27.1 | 35 | 72.9 | |||
Depression | |||||||
Has decreased or stayed the same | 33 | 70.2 | 14 | 29.8 | 58.802 | 2 | < 0.001 |
Has increased | 31 | 83.8 | 6 | 16.2 | |||
Never had it | 94 | 30.5 | 214 | 69.5 | |||
Anxiety | |||||||
Has decreased or stayed the same | 36 | 46.8 | 41 | 53.2 | 71.787 | 2 | < 0.001 |
Has increased | 53 | 85.5 | 9 | 14.5 | |||
Never had it | 69 | 27.3 | 184 | 72.7 | |||
Stress | |||||||
Has decreased or stayed the same | 45 | 44.6 | 56 | 55.4 | 69.822 | 2 | < 0.001 |
Has increased | 60 | 78.9 | 16 | 21.1 | |||
Never had it | 53 | 24.7 | 162 | 75.3 | |||
Alcohol | |||||||
Never used | 81 | 33.2 | 163 | 66.8 | 14.963 | 2 | 0.001 |
Reduced | 27 | 49.1 | 28 | 50.9 | |||
Used the same or increased | 50 | 54.9 | 41 | 45.1 | |||
Cigarettes, e-cigarettes, vapes | |||||||
Never used | 120 | 37.2 | 203 | 62.8 | 9.148 | 2 | 0.010 |
Reduced | 14 | 46.7 | 16 | 53.3 | |||
Used the same or increased | 24 | 61.5 | 15 | 38.5 | |||
Cannabis | |||||||
Never used | 126 | 37.1 | 214 | 62.9 | 12.305 | 2 | 0.002 |
Reduced | 15 | 60.0 | 10 | 40.0 | |||
Used the same or increased | 17 | 65.4 | 9 | 34.6 | |||
Other drugs | |||||||
Never used | 136 | 38.3 | 219 | 61.7 | 8.499 | 2 | 0.014 |
Reduced | 13 | 54.2 | 11 | 45.8 | |||
Used the same or increased | 9 | 75.0 | 3 | 25.0 | |||
Univariate analysis was conducted using chi-square test and t-test. p < 0.05 was considered significant. |
df, degrees of freedom; SD, standard deviation; UK, United Kingdom.
Higher levels of anxiety were found in Years 11 and higher (51.1% vs 25.7%, p < 0.001), females (50.3% vs 32.6%, p = 0.001), residents of the UK and other countries versus Thailand (49.0%, 48.1% vs 35.3%, p = 0.031) (Table 4.1) and older participants (M = 15.97, SD = 1.56 for anxious vs M = 15.24, SD = 1.75 for not, p = 0.005) (Table 3).
Anxiety was higher in those who reported previous depression, anxiety, and stress (p < 0.001). Higher levels of anxiety were also reported in those who reported not using public transport since school closure, and ‘sometimes or regularly’ had sleep problems due to thinking about COVID-19 (p < 0.05). In addition, those who perceived problems with the consequences of social distancing, including online learning, inability to eat out and inability to go shopping, all displayed higher levels of anxiety (p < 0.05). Conversely, the lowest levels of anxiety were found in those who were the most worried about going outside (p = 0.004) (Table 3).
Those who reported use of alcohol, cigarettes/e-cigarettes/vapes, cannabis, and/or other drugs experienced higher levels of anxiety than those who never used them. Additionally, those who experienced unchanged or increased use of all four substance categories during lockdown experienced higher levels of anxiety than those who experienced reduced use (p < 0.05) (Table 3).
We found that females (compared to males, OR = 3.075, 95% CI = 1.60–5.92) and participants who had not used public transport at all since the physical closure of their school (compared to those who sometimes used it, OR = 8.31, 95% CI = 2.49-27.755) had significantly increased risk of anxiety (Table 4).
Variables | Exhibiting anxiety symptoms | ||
---|---|---|---|
ORs | 95% CI | p-values | |
Gender | |||
Female | 3.075 | 1.60–5.92 | 0.001 |
Male | 1 | ||
Using public transport since the physical closure of school | |||
Not at all/No | 8.31 | 2.49–27.755 | 0.001 |
Sometimes/Yes | 1 | ||
Depression | |||
Has decreased or stayed the same | 6.585 | 2.13–20.30 | 0.001 |
Has increased | 3.21 | 0.88–11.63 | 0.076 |
Never had it | 1 | ||
Anxiety | |||
Has decreased or stayed the same | 1.08 | 0.42–2.80 | 0.876 |
Has increased | 4.61 | 1.54–13.795 | 0.006 |
Never had it | 1 | ||
Stress | |||
Has decreased or stayed the same | 1.13 | 0.50–2.54 | 0.774 |
Has increased | 3.89 | 1.52–9.92 | 0.005 |
Never had it | 1 | ||
Multivariate analysis was conducted using logistic nominal regression. p < 0.05 was considered significant. |
OR, odds ratio; CI, confidence interval.
Higher levels of anxiety were found to be significantly associated with an increase in levels of previous anxiety (OR = 4.61, 95% CI = 1.54-13.795), and previous stress (OR = 3.89, 95% CI = 1.52–9.92) during lockdown, compared to those who reported never having previous anxiety or stress. Those who reported decreased or equal levels of previous depression during lockdown still had higher levels of anxiety than those who never had previous depression before the lockdown (OR = 6.585, 95% CI = 2.14–20.30).
Factors associated with perceived stress
Variables | High levels of perceived stress | Low/moderate levels of perceived stress | |||||
---|---|---|---|---|---|---|---|
n | % | n | % | χ2 | df | p-value | |
UK School Year | |||||||
8/9/10 | 5 | 3.0 | 162 | 97.0 | 14.919 | 1 | < 0.001 |
11/12/13 | 33 | 14.7 | 192 | 85.3 | |||
Female | 29 | 17.2 | 140 | 82.8 | 29.323 | 2 | < 0.001 |
Age (mean/SD) | (16.5) | (1.1) | (15.4) | (1.7) | t388= -3.726 | - | < 0.001 |
Country currently living in | |||||||
Thailand | 21 | 9.1 | 211 | 90.9 | 2.682 | 2 | 0.262 |
UK | 14 | 13.5 | 90 | 86.5 | |||
Others | 3 | 5.8 | 49 | 94.2 | |||
Country currently studying in | |||||||
Thailand | 15 | 7.8 | 178 | 92.2 | 4.384 | 2 | 0.112 |
UK | 22 | 13.2 | 145 | 86.8 | |||
Others | 1 | 3.4 | 28 | 96.6 | |||
Area currently living in | |||||||
Urban | 12 | 20.3 | 47 | 79.7 | 8.893 | 2 | 0.012 |
Peri-urban | 7 | 8.2 | 78 | 91.8 | |||
Rural | 19 | 7.7 | 227 | 92.3 | |||
Being under mandatory self-quarantine | 25 | 12.8 | 170 | 87.2 | 4.333 | 1 | 0.037 |
Level of feeling that COVID-19 affects daily life | |||||||
Great effect | 26 | 12.0 | 191 | 88.0 | 3.100 | 2 | 0.212 |
Slight/little effect | 10 | 6.5 | 144 | 93.5 | |||
No effect | 2 | 9.5 | 19 | 90.5 | |||
Using public transport since the physical closure of school | 1 | 1.5 | 66 | 98.5 | 6.209 | 1 | 0.013 |
Level of thinking of catching COVID-19 | |||||||
High chance | 2 | 9.5 | 19 | 90.5 | 0.696 | 2 | 0.706 |
Moderate chance | 11 | 8.0 | 126 | 92.0 | |||
Little/no chance | 25 | 10.7 | 209 | 89.3 | |||
Being worried about going outside | |||||||
Not at all | 11 | 8.6 | 117 | 91.4 | 0.292 | 2 | 0.864 |
Slightly/a bit worried | 20 | 10.4 | 172 | 89.6 | |||
Very worried | 7 | 9.7 | 65 | 90.3 | |||
Problems with sleep by thinking about COVID-19? | 17 | 18.7 | 74 | 81.3 | 10.935 | 1 | 0.001 |
Online learning | |||||||
Experience with little or many problems | 32 | 12.9 | 216 | 87.1 | 9.295 | 2 | 0.010 |
Experience without any problems | 3 | 2.7 | 110 | 97.3 | |||
No experience | 3 | 10.0 | 27 | 90.0 | |||
Cannot meet friends regularly | |||||||
Experience with problems | 35 | 11.6 | 267 | 88.4 | 5.626 | 2 | 0.060 |
Experience without any problems | 3 | 4.3 | 66 | 95.7 | |||
No experience | 0 | 0.0 | 20 | 100.0 | |||
Cannot go out to eat | |||||||
Experience with problems | 27 | 12.2 | 194 | 87.8 | 4.304 | 2 | 0.116 |
Experience without any problems | 8 | 5.6 | 134 | 94.4 | |||
No experience | 3 | 10.7 | 25 | 89.3 | |||
Cannot go out to shop | |||||||
Experience with problems | 23 | 12.6 | 159 | 87.4 | 3.367 | 2 | 0.186 |
Experience without any problems | 12 | 7.5 | 149 | 92.5 | |||
No experience | 3 | 6.3 | 45 | 93.8 | |||
Depression | |||||||
Has decreased or stayed the same | 4 | 8.5 | 43 | 91.5 | 30.341 | 2 | < 0.001 |
Has increased | 13 | 35.1 | 24 | 64.9 | |||
Never had it | 21 | 6.8 | 287 | 93.2 | |||
Anxiety | |||||||
Has decreased or stayed the same | 8 | 10.4 | 69 | 89.6 | 23.431 | 2 | < 0.001 |
Has increased | 16 | 25.8 | 46 | 74.2 | |||
Never had it | 14 | 5.5 | 239 | 94.5 | |||
Stress | |||||||
Has decreased or stayed the same | 8 | 7.9 | 93 | 92.1 | 41.307 | 2 | < 0.001 |
Has increased | 22 | 28.9 | 54 | 71.1 | |||
Never had it | 8 | 3.7 | 207 | 96.3 | |||
Alcohol | |||||||
Never used | 19 | 7.8 | 225 | 92.2 | 3.282 | 2 | 0.194 |
Reduced | 6 | 10.9 | 49 | 89.1 | |||
Used the same or increased | 13 | 14.3 | 78 | 85.7 | |||
Cigarettes, e-cigarettes, vapes | |||||||
Never used | 26 | 8.0 | 297 | 92.0 | 6.666 | 2 | 0.036 |
Reduced | 4 | 13.3 | 26 | 86.7 | |||
Used the same or increased | 8 | 20.5 | 31 | 79.5 | |||
Cannabis | |||||||
Never used | 27 | 7.9 | 313 | 92.1 | 11.119 | 2 | 0.004 |
Reduced | 4 | 16.0 | 21 | 84.0 | |||
Used the same or increased | 7 | 26.9 | 19 | 73.1 | |||
Other drugs | |||||||
Never used | 32 | 9.0 | 323 | 91.0 | 2.182 | 2 | 0.336 |
Reduced | 4 | 16.7 | 20 | 83.3 | |||
Used the same or increased | 2 | 16.7 | 10 | 83.3 | |||
Univariate analysis was conducted using chi-square test and t-test. p < 0.05 was considered significant. |
df, degrees of freedom; SD, standard deviation; UK, United Kingdom.
Higher levels of high stress were found in Years 11 and higher (14.7% vs 3.0%, p < 0.001), females (17.2% vs 3.6%, p < 0.001) (Table 6.1) and older participants (M = 16.50, SD = 1.08 for high stress vs M = 15.24, SD = 1.75 for low/moderate stress, p < 0.001) (Table 6.2).
High stress levels were greater in those who reported previous depression, anxiety, and stress which had increased/worsened due to lockdown (p < 0.001). Higher levels of high stress were also reported in those who lived in rural areas, reported not using public transport since school closure, ‘sometimes or regularly’ had sleep problems due to thinking about COVID-19, had undergone mandatory self-quarantine, and perceived problems with online learning (p < 0.05) (Table 6.1).
Those who reported use of cigarettes/e-cigarettes/vapes and cannabis experienced higher levels of high stress than those who never used them. Additionally, those who experienced unchanged or increased use of the two substances experienced higher levels of high stress than those who experienced reduced use (p < 0.05) (Table 6.1).
Significantly increased risk of high stress was found in those who had no worry (OR = 2.705, 95% CI = 1.02–7.205) or slight worry (OR = 2.78, 95% CI = 1.19–6.53) about going outside, compared to those very worried. Those who found not being able to go shopping problematic (compared to those who were able to go shopping, OR = 4.03, 95% CI = 1.24–13.06) had significantly increased risk of high stress. Participants who experienced increased or unchanged alcohol use during lockdown were found to not have as high levels of high stress as those who had never used alcohol (OR = 0.27, 95% CI = 0.11–0.69) (Table 6).
Variables | Higher levels of high stress | ||
---|---|---|---|
ORs | 95% CI | p-values | |
Depression | |||
Has decreased or stayed the same | 14.61 | 2.83–75.42 | 0.001 |
Has increased | 0.66 | 0.14–3.02 | 0.590 |
Never had it | 1 | ||
Anxiety | |||
Has decreased or stayed the same | 0.52 | 0.17–1.56 | 0.242 |
Has increased | 4.53 | 0.14–3.02 | 0.045 |
Never had it | 1 | ||
Stress | |||
Has decreased or stayed the same | 2.08 | 0.845–5.14 | 0.111 |
Has increased | 15.43 | 3.665–64.98 | 0.000 |
Never had it | 1 | ||
Being worried about going outside | |||
Not at all | 2.705 | 1.02–7.205 | 0.046 |
Slightly/a bit worried | 2.78 | 1.19–6.53 | 0.019 |
Very worried | 1 | ||
Cannot go out to shop | |||
Experience with little or many problems | 4.03 | 1.24–13.06 | 0.02 |
Experience without any problems | 1.49 | 0.49–4.54 | 0.48 |
No experience | 1 | ||
Alcohol | |||
Used the same or increased | 0.27 | 0.11–0.69 | 0.006 |
Reduced | 0.53 | 0.18–1.53 | 0.241 |
Never used | 1 | ||
Multivariate analysis was conducted using logistic nominal regression. p < 0.05 was considered significant. |
OR, odds ratio; CI, confidence interval.
Higher levels of high stress were found to be significantly associated with an increase in levels of previous anxiety (OR = 4.53, 95% CI = 1.03–19.83), and previous stress (OR = 15.43, 95% CI = 3.665–64.98), compared to those who never had anxiety or stress during lockdown. Those who reported decreased or equal levels of previous depression during lockdown still had higher levels of high stress than those who never had previous depression before the lockdown (OR = 14.61, 95% CI = 2.83–75.42).
We conducted an online survey study which found a high prevalence of mental health conditions among secondary school students across a range of countries during COVID-19 lockdown in May and June 2020. Approximately 60% had depression, 40% had anxiety and 10% experienced a high level of stress. Being female enhanced depression and anxiety. Being enrolled in the last three years of high/secondary school enhanced depression. Changes in patterns of substance use were also significantly associated with anxiety and stress.
Although one must take into account the environment in which adolescents live in, as it will ultimately affect their mental health, it is clear to see that our findings demonstrate a higher prevalence of depressive, anxiety and stress symptoms. Compared to previous studies done during arguably more normal times, our prevalence of depressive symptoms was higher than 53.2% in Norwegian secondary school students (20), 55.9% in Nigerian secondary school students (21), 52.9% in Chinese adolescents (22). Similarly, our findings indicate a higher prevalence of anxiety compared to around 10% in Canadian secondary school students (23) and higher-education students from the UK (24) from studies conducted during non-COVID19 times. High stress level was also found to be more common than the 4% among Thai students aged 15–19 years (25) before COVID-19.
Furthermore, a recent study conducted on Chinese adolescents aged 12–18 years also demonstrated similar findings of elevated prevalence (43.7% exhibiting depressive symptoms, 37.4% exhibiting symptoms of anxiety) during COVID-19 (15). There were no reports of stress prevalence among secondary school students during COVID-19.
In our study, being female was found to be significantly associated with higher levels of depression and anxiety. This was found to be in accordance with previous studies concerning lockdown mental health (9, 26, 27), as well as during more general times (28, 29). However, Cao et al. did not find gender to be significantly associative with higher levels of anxiety during COVID-19 lockdown (30).
Being enrolled in senior high school (Years 11–13 or Grades 10–12) was found to be significantly associated with higher levels of depression, but not anxiety, in our study. Zhou et al. found similarly significant associations with both anxiety and depression during COVID-19 (15).
A combination of emotions such as boredom, anxiety, depression and fear may lead to increased substance use as a means of coping (31). With many countries closing down shops and public services (32), with some even enacting coronavirus alcohol bans (33), many may find themselves in forced abstinence (12, 34). Access to recreational substances normally used which have been limited, for example due to lockdown or otherwise, can exacerbate mental health effects (10, 11, 35). We found that reports of decreased cannabis use were significantly associated with higher levels of depression. Furthermore, increased or unchanged alcohol use was found to be associated with lower levels of stress. Despite likely being a protective factor in this study, adolescent alcohol use has been found to predict development of alcohol problems into young adulthood (36), so should be treated with caution. In times of lockdown, those with substance use disorders tend to be ignored (10). This problem is further worsened due to the fact that adolescents who use substances tend to not want to disclose their use, for example to their parents. Therefore, there is a chance that those concealing their use within their household may face adverse effects. Lack of access to supervised substance use may also increase hazardous use, due to interruption of opening hours of harm reduction services as well as general fears of COVID-19 infection among clients (37). The United Nations Office on Drugs and Crime (UNODC) reported in May 2020 that COVID-19 may have adverse effects on drug supply chains. This may lead to harmful adaptations by both producers, such as reduced street purity, and users, such as a shift towards drug injection as well as sharing paraphernalia (38).
There are some limitations in this study. This is a cross-sectional study and we were therefore not able to assess any long-term impacts or progressions of mental health conditions like a longitudinal study would be able to. Furthermore, due to our use of snowball sampling, we were unable to record or know the number of people who had been formally ‘invited’ to take part in the survey. Moreover, we found that other studies examining the psychological impacts of lockdown, particularly concerning past outbreaks, assessed post-traumatic stress disorder (PTSD) as well. We did not choose to assess this as the progression of the COVID-19 pandemic is far from over. Different countries will also have been affected with varying severity, which could ultimately lead to different PTSD outcomes. Almost 60% of our participants lived in Thailand, where COVID-19 has struck relatively less hard. As of 7 September 2020, Thailand has had 3,445 COVID-19 cases, compared to 347,152 cases in the UK. Their death rates per million people are also extremely different, with 0.8 for Thailand, but 611 for the UK (18). Nevertheless, it will be important to assess it once lockdown ends, as studies have demonstrated the long-lasting PTSD caused by lockdown and quarantine (31).
In conclusion, amidst lockdown measures, around half of adolescents exhibited depression and/or anxiety. Particular attention must be paid to females, older adolescents and substance users. Girl-centred mental health support platforms (39, 40) should be readily available, and tailored to fit specific countries’ contexts. Schools must closely monitor and act upon any concerns which arise from their students, especially the more senior individuals who may be particularly stressed about academic commitments such as exam cancellations (13), university preparation and work experience opportunities. However, apart from their parents, schools must also monitor the child’s mental health wellbeing in general as school forms a very important aspect of their life and these changes due to COVID-19 could be drastic for some. Harm reduction services must receive support to maintain service delivery, perhaps adapting and utilising innovative interventions such as telemedicine-delivered prescription and treatment (41), as well as a focus on more tailoring towards adolescent users.
Ethics approval and consent to participate
The study was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University. Consent by action was obtained via a digitalised consent form, explaining the objectives and contents of the survey as well as potential risks and intended benefits. Despite most participants being under 18, parental consent was waived due to the anonymous nature of the survey.
Consent for publication
Not applicable
Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. All data generated or analysed during this study are included in this published article.
Competing interests
The authors declare that they have no competing interests.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Authors' contributions
GP conceptualized and designed the study, collected data, performed formal analysis, wrote the first draft of manuscript, reviewed and edited the manuscript. RK designed the study, supervised data collection, performed formal analysis, reviewed and edited the manuscript. Both authors read and approved the final manuscript.
Acknowledgments
We would like to thank Thitima Duangsanit for facilitating the logistics in conducting this study during the COVID-19 outbreak. We would also like to thank all participants for their time and efforts.